{
"best_metric": null,
"best_model_checkpoint": null,
"epoch": 3.0,
"eval_steps": 500,
"global_step": 4323,
"is_hyper_param_search": false,
"is_local_process_zero": true,
"is_world_process_zero": true,
"log_history": [
{
"epoch": 0.03,
"learning_rate": 7.692307692307694e-06,
"loss": 3.6444,
"step": 50
},
{
"epoch": 0.07,
"learning_rate": 1.5384615384615387e-05,
"loss": 3.0454,
"step": 100
},
{
"epoch": 0.1,
"learning_rate": 1.9998877278887434e-05,
"loss": 2.7656,
"step": 150
},
{
"epoch": 0.14,
"learning_rate": 1.9986249561409415e-05,
"loss": 2.5791,
"step": 200
},
{
"epoch": 0.17,
"learning_rate": 1.9959608504164855e-05,
"loss": 2.4997,
"step": 250
},
{
"epoch": 0.21,
"learning_rate": 1.9918991491577664e-05,
"loss": 2.4208,
"step": 300
},
{
"epoch": 0.24,
"learning_rate": 1.9864455520020126e-05,
"loss": 2.3673,
"step": 350
},
{
"epoch": 0.28,
"learning_rate": 1.9796077117831955e-05,
"loss": 2.3077,
"step": 400
},
{
"epoch": 0.31,
"learning_rate": 1.971395223793088e-05,
"loss": 2.2695,
"step": 450
},
{
"epoch": 0.35,
"learning_rate": 1.9618196123165404e-05,
"loss": 2.2326,
"step": 500
},
{
"epoch": 0.38,
"learning_rate": 1.9508943144598726e-05,
"loss": 2.2122,
"step": 550
},
{
"epoch": 0.42,
"learning_rate": 1.9386346612950727e-05,
"loss": 2.1661,
"step": 600
},
{
"epoch": 0.45,
"learning_rate": 1.925057856346263e-05,
"loss": 2.1537,
"step": 650
},
{
"epoch": 0.49,
"learning_rate": 1.9101829514486206e-05,
"loss": 2.1359,
"step": 700
},
{
"epoch": 0.52,
"learning_rate": 1.8940308200136353e-05,
"loss": 2.0954,
"step": 750
},
{
"epoch": 0.56,
"learning_rate": 1.8766241277382096e-05,
"loss": 2.0736,
"step": 800
},
{
"epoch": 0.59,
"learning_rate": 1.8579873007987135e-05,
"loss": 2.0633,
"step": 850
},
{
"epoch": 0.62,
"learning_rate": 1.838146491574624e-05,
"loss": 2.0746,
"step": 900
},
{
"epoch": 0.66,
"learning_rate": 1.8171295419498417e-05,
"loss": 2.0407,
"step": 950
},
{
"epoch": 0.69,
"learning_rate": 1.794965944243194e-05,
"loss": 2.0361,
"step": 1000
},
{
"epoch": 0.73,
"learning_rate": 1.7716867998229368e-05,
"loss": 2.0215,
"step": 1050
},
{
"epoch": 0.76,
"learning_rate": 1.747324775463342e-05,
"loss": 1.9921,
"step": 1100
},
{
"epoch": 0.8,
"learning_rate": 1.721914057504601e-05,
"loss": 2.0012,
"step": 1150
},
{
"epoch": 0.83,
"learning_rate": 1.695490303880384e-05,
"loss": 1.9912,
"step": 1200
},
{
"epoch": 0.87,
"learning_rate": 1.6680905940803596e-05,
"loss": 1.9723,
"step": 1250
},
{
"epoch": 0.9,
"learning_rate": 1.6397533771179006e-05,
"loss": 1.9712,
"step": 1300
},
{
"epoch": 0.94,
"learning_rate": 1.6105184175759864e-05,
"loss": 1.9479,
"step": 1350
},
{
"epoch": 0.97,
"learning_rate": 1.5804267398070123e-05,
"loss": 1.9513,
"step": 1400
},
{
"epoch": 1.01,
"learning_rate": 1.549520570364815e-05,
"loss": 1.9449,
"step": 1450
},
{
"epoch": 1.04,
"learning_rate": 1.5178432787496863e-05,
"loss": 1.9101,
"step": 1500
},
{
"epoch": 1.08,
"learning_rate": 1.4854393165495381e-05,
"loss": 1.9104,
"step": 1550
},
{
"epoch": 1.11,
"learning_rate": 1.4523541550626093e-05,
"loss": 1.8977,
"step": 1600
},
{
"epoch": 1.15,
"learning_rate": 1.4186342214892513e-05,
"loss": 1.9062,
"step": 1650
},
{
"epoch": 1.18,
"learning_rate": 1.3843268337823312e-05,
"loss": 1.9024,
"step": 1700
},
{
"epoch": 1.21,
"learning_rate": 1.3494801342476752e-05,
"loss": 1.9043,
"step": 1750
},
{
"epoch": 1.25,
"learning_rate": 1.3141430219877257e-05,
"loss": 1.8892,
"step": 1800
},
{
"epoch": 1.28,
"learning_rate": 1.2783650842832161e-05,
"loss": 1.884,
"step": 1850
},
{
"epoch": 1.32,
"learning_rate": 1.2421965270091454e-05,
"loss": 1.8725,
"step": 1900
},
{
"epoch": 1.35,
"learning_rate": 1.205688104182709e-05,
"loss": 1.8847,
"step": 1950
},
{
"epoch": 1.39,
"learning_rate": 1.168891046742033e-05,
"loss": 1.8765,
"step": 2000
},
{
"epoch": 1.42,
"learning_rate": 1.1318569906556733e-05,
"loss": 1.8787,
"step": 2050
},
{
"epoch": 1.46,
"learning_rate": 1.0946379044637374e-05,
"loss": 1.8805,
"step": 2100
},
{
"epoch": 1.49,
"learning_rate": 1.0572860163523335e-05,
"loss": 1.8752,
"step": 2150
},
{
"epoch": 1.53,
"learning_rate": 1.0198537408636583e-05,
"loss": 1.8843,
"step": 2200
},
{
"epoch": 1.56,
"learning_rate": 9.823936053445839e-06,
"loss": 1.8582,
"step": 2250
},
{
"epoch": 1.6,
"learning_rate": 9.449581762369454e-06,
"loss": 1.8684,
"step": 2300
},
{
"epoch": 1.63,
"learning_rate": 9.075999853129733e-06,
"loss": 1.8533,
"step": 2350
},
{
"epoch": 1.67,
"learning_rate": 8.703714559593719e-06,
"loss": 1.8464,
"step": 2400
},
{
"epoch": 1.7,
"learning_rate": 8.333248296134922e-06,
"loss": 1.8423,
"step": 2450
},
{
"epoch": 1.73,
"learning_rate": 7.965120924548302e-06,
"loss": 1.8669,
"step": 2500
},
{
"epoch": 1.77,
"learning_rate": 7.599849024547167e-06,
"loss": 1.8493,
"step": 2550
},
{
"epoch": 1.8,
"learning_rate": 7.237945168865693e-06,
"loss": 1.8523,
"step": 2600
},
{
"epoch": 1.84,
"learning_rate": 6.879917203984306e-06,
"loss": 1.8661,
"step": 2650
},
{
"epoch": 1.87,
"learning_rate": 6.526267537487263e-06,
"loss": 1.8445,
"step": 2700
},
{
"epoch": 1.91,
"learning_rate": 6.177492433052374e-06,
"loss": 1.8412,
"step": 2750
},
{
"epoch": 1.94,
"learning_rate": 5.834081314062314e-06,
"loss": 1.8391,
"step": 2800
},
{
"epoch": 1.98,
"learning_rate": 5.496516076814623e-06,
"loss": 1.8479,
"step": 2850
},
{
"epoch": 2.01,
"learning_rate": 5.165270414294229e-06,
"loss": 1.8553,
"step": 2900
},
{
"epoch": 2.05,
"learning_rate": 4.8408091514573454e-06,
"loss": 1.8322,
"step": 2950
},
{
"epoch": 2.08,
"learning_rate": 4.523587592959557e-06,
"loss": 1.8343,
"step": 3000
},
{
"epoch": 2.12,
"learning_rate": 4.214050884243429e-06,
"loss": 1.8277,
"step": 3050
},
{
"epoch": 2.15,
"learning_rate": 3.912633386882113e-06,
"loss": 1.8368,
"step": 3100
},
{
"epoch": 2.19,
"learning_rate": 3.6197580690556233e-06,
"loss": 1.819,
"step": 3150
},
{
"epoch": 2.22,
"learning_rate": 3.335835912015e-06,
"loss": 1.8237,
"step": 3200
},
{
"epoch": 2.26,
"learning_rate": 3.0612653333673092e-06,
"loss": 1.8163,
"step": 3250
},
{
"epoch": 2.29,
"learning_rate": 2.7964316279907555e-06,
"loss": 1.8223,
"step": 3300
},
{
"epoch": 2.32,
"learning_rate": 2.541706427364431e-06,
"loss": 1.8375,
"step": 3350
},
{
"epoch": 2.36,
"learning_rate": 2.2974471780714147e-06,
"loss": 1.8157,
"step": 3400
},
{
"epoch": 2.39,
"learning_rate": 2.0639966402070154e-06,
"loss": 1.8172,
"step": 3450
},
{
"epoch": 2.43,
"learning_rate": 1.8416824063960336e-06,
"loss": 1.7918,
"step": 3500
},
{
"epoch": 2.46,
"learning_rate": 1.6308164420939909e-06,
"loss": 1.8149,
"step": 3550
},
{
"epoch": 2.5,
"learning_rate": 1.4316946478173732e-06,
"loss": 1.822,
"step": 3600
},
{
"epoch": 2.53,
"learning_rate": 1.2445964439172277e-06,
"loss": 1.8282,
"step": 3650
},
{
"epoch": 2.57,
"learning_rate": 1.069784378478781e-06,
"loss": 1.819,
"step": 3700
},
{
"epoch": 2.6,
"learning_rate": 9.075037588972901e-07,
"loss": 1.8158,
"step": 3750
},
{
"epoch": 2.64,
"learning_rate": 7.579823076471216e-07,
"loss": 1.8106,
"step": 3800
},
{
"epoch": 2.67,
"learning_rate": 6.214298427270948e-07,
"loss": 1.8179,
"step": 3850
},
{
"epoch": 2.71,
"learning_rate": 4.980379832305604e-07,
"loss": 1.8284,
"step": 3900
},
{
"epoch": 2.74,
"learning_rate": 3.8797988045329635e-07,
"loss": 1.8309,
"step": 3950
},
{
"epoch": 2.78,
"learning_rate": 2.914099749166244e-07,
"loss": 1.8219,
"step": 4000
},
{
"epoch": 2.81,
"learning_rate": 2.08463779646646e-07,
"loss": 1.8182,
"step": 4050
},
{
"epoch": 2.85,
"learning_rate": 1.3925769001376876e-07,
"loss": 1.8142,
"step": 4100
},
{
"epoch": 2.88,
"learning_rate": 8.388882039931934e-08,
"loss": 1.8152,
"step": 4150
},
{
"epoch": 2.91,
"learning_rate": 4.243486791847829e-08,
"loss": 1.8172,
"step": 4200
},
{
"epoch": 2.95,
"learning_rate": 1.495400339075359e-08,
"loss": 1.8079,
"step": 4250
},
{
"epoch": 2.98,
"learning_rate": 1.4847897109848685e-09,
"loss": 1.8273,
"step": 4300
},
{
"epoch": 3.0,
"step": 4323,
"total_flos": 1.477332496684155e+17,
"train_loss": 1.9787378639858297,
"train_runtime": 6196.198,
"train_samples_per_second": 44.65,
"train_steps_per_second": 0.698
}
],
"logging_steps": 50,
"max_steps": 4323,
"num_input_tokens_seen": 0,
"num_train_epochs": 3,
"save_steps": 1200,
"total_flos": 1.477332496684155e+17,
"train_batch_size": 8,
"trial_name": null,
"trial_params": null
}